In a multi agent or stochastic environment it's hard to select one action for an agent. Many papers are related for this task but aren't implementable in the real world.
Hi,
What do you mean by "implementable in the real world". Is there no implementation in these papers ?
Usually, in this type of process, you have, for each agent, a bunch of actions to do. Each action has also often a probability to appear (to be executed) according to your model. So, your simulation will select it stochasticly according to the probabilities.
Well, that's hard to answer without knowing about the specifics of the problem. My advice would be to have a look at the different models for modelling multi-agent sequential decision making under uncertainty, and see what fits your problem best. There are tutorials on the webpage of this year's MSDM (Multi-Agent Sequential Decision Making under uncertainty) workshop, that contain a pretty nice overview.
http://masplan.org/msdm2014:tutorial
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Artificial Intelligence